Uncertainty Estimation Based on Repeated Observations of a Population of Instruments

by Jonathon Harben

This paper explores the similarities and differences between a measurement uncertainty as described in the GUM and a manufacturer’s specification when the history of compliance testing results is taken into consideration. It has been suggested in certain cases where compliance to specifications is asserted, it can be sufficient to report only the historical reliability which encompasses all uncertainties as well as the intrinsic variability of the process. Consideration of metrological traceability is addressed, where verification of compliance with the stated specification is asserted with only the knowledge of the historical reliability as the measurement uncertainty.